俱舍論疏 5.263

5.263

論「毘婆沙師至所能了故」，此迷有部宗也。《正理論》云「《毘婆沙》說名、句、文三各有三種。名三種者，謂名、名身、多名身；句、文亦爾。名有多位，謂一字生、或二字生、或多字生。一字生者，說一字時但可有名，說二字時即謂名身。或作是說：說三字時即謂多名身。或作是說：說四字時方謂多名身。二字生者，說二字時但可有名，說四字時即謂名身。或作是說：說六字時即謂多名身。或作是說：說八字時方謂多名身。多字生中三字生者，說三字時但可有名，說六字時即謂名身。或作是說：說九字時即謂多名身。或作是說：說十二字時方謂多名身。此為門故，餘多字生名身、多身，如理應說。句亦多位，謂處中句、初句後句、短句長句。若八字生名處中句，不長不短故謂處中。三十二字生於四句，如是四句成室路迦，經、論文章多依此數。若六字已上生名初句，二十六字以下生名後句。若減六字生名短句，過二十六字生名長句。且依處中句辨三種：說八字時但可有句，說十六字時即謂句身。或作是說：說二十四字時即謂多句身。或作是說：說三十二字時方謂多句身。文即字故唯有一位，說一字時但可有文，說二字時即謂文身。或作是說：說三字時即謂多文身。或作是說：說四字時方謂多文身。由此理故應作是說：說一字時有名、無名身、無多名身；無句、無句身、無多句身；有文、無文身、無多文身。說二字時，有名、有名身、無多名身；無句等三；有文、有文身、無多文身。說四字時，有名等三，無句等三，有文等三。說八字時，有名等三；有句、無句身、無多句身；有文等三。說十六字時，有名等三，有句、有句身、無多句身；有文等三。說三十二字時，名、句、文三各具三種。由此為門，餘如理說。」准上論文，多念聲生一文，前念聲無文。一文生一名及一字句者，即一念聲三法同起，謂文、名、句。若二字名已去，於前念文未有名、句，至其後字成名、句位方有名、句。多文生一名、多名成一句者，非一一念聲皆有文、一一文皆有名、一一名皆有句，然有一字生名，容有一字一名、一句，或無一字句，句必須二字。其二字已上生名者，皆於第二字等方有二字名生。若二字乃至多字生名者，於前字即得生解。此由聞者串習於前解義，非謂於前位中有二字名等於先已生。如言「諸行」二字名，於中有二：一二一字名、二一二字名。餘例准此。所以如是如說二字名時，一人聞諸、一人聞行，各生一解故。若聞一兼二，即是串習。若作此釋，善順名之極少，謂一字也。又如說「諸行無常」等句，諸行是自性、無常是差別，名詮自性、句詮差別。無常二字，或有名起、或無名起，義則未定。如說無常可厭，無常即是自性、可厭即是差別。此無常二字理合有名，若為差別即無名起。如實說者，若為自性、若為差別皆悉有名。如人說諸行無常句，有人唯聞無字、或唯聞常字、或聞無常二字、或具聞四字。若唯聞無字者，謂唯說無。若聞常字者，謂唯說常。若合聞者，謂詮無常。然此無常未屬何法，故知此上所聞皆不了句。若有聞說諸行無常，此方了句，知此無常諸行差別故。以此准知，諸行無常四字之邊，一一皆有一字名也。於中若至二字，亦有二字名也。至第三字，有一字名及三字句。如有但聞「諸行無」三字，此即謂詮諸行無也。諸行為自性、無為差別，至第四字有一字名，及二字名，及三字名，及三、四字句也。於中加減如理應思。又釋所以四字名等有一字名，發語遠加行心多緣此故。如說「諸行」兩字，必須先緣詮諸詮行一一別名，方可合成二字名故。《正理》、《婆沙》說名、句、文各有三種：一名、二名身、三多名身；句、文亦爾。名有多位，有一字生、有二字生、有多字生。《婆沙》名有六位，有一字、二字、四字、八字、十六字、三十二字。准此論，據位相倍及極一頌，作如是說。然實以論非無三字、五字、七字、九字等生名之者，上引論云一字生名、二字生名、多字生名。多字生名者，或三字已上生名，皆多字名。以西方語法，說二名身有合集義故，說三已上皆名多故。或說四字已上名多字生也，倍二字故。一字生名者，說一字時但名名，說二字時有名身也。多名身有二釋，或作是說：說三字時即謂多名身；或作是說：說四字時即謂多名身。有人云：問多名身中，一云說三字時名多名身，一云說四字時名多名身，其義云何？解云：三字生名名多名身者，第一、第二字為一身，第一、第三字復為一身，第二、第三字復為一身。依聲明法，言三已去方說為多。若四字生名名多名身者，隣次相合即為三身，名多名身。又解：但約名多故名多名身，不約多身。又解：身上加身名多名身。准上三釋自相牟楯，但遍說而已，不能是非。何者？准論所明，但說名、句、文三各有三位，謂一、二、多。一以名一故不成身，二以名二故成身也，第三亦合以名多故名多名身。因何說一二之位即是名之一二？第三多位因何其數即說是身？若於此中以身多故名多名身，即是第三說其身數，於中亦合說身及身身。依第二解，四字生多隣次相對應名身身。又自引聲明說三已上方得名多，因何復說身上加身名為多身？若爾，三字生者次第相合亦有二身，是身上加身，應亦得名多名身也。雖一解但約名多故名多名身，既不得所由，即是如虫食木。此中所說一、二、多三皆約名說，不據身論。《婆沙》更有明文，應撿錄之。又牒文云三字生名名多名身者，謬。三字生名者，唯有一三字名故。此中兩說不同。一據成多即名多身，謂三已上。二據位相倍，謂四名已上倍於前也。兼依二釋，義皆無違，故論雙舉。然唯有字、名、句體而無頌體者，其字、名、句，五趣諸方愚智不同，但能語者皆聲發文，依文有名，依名有句，皆能說自性、差別故，皆令解自性、差別。故其頌唯是文者能作，只是安布其文乃至名、句，更無別物。問：如說一名、一字，對諸方名各生異解，為有多名、為有一名？為有一字、為有多字？答：字唯是一，諸方通用更無差別。名即不同，若為此方語，唯有此方名，無諸方名。所以知然？若不爾者，即應唯說一字生名應有多名。論既不說多名，故知唯有一名。又當發語遠方便心唯緣一方名故。問：若爾，何故生其多解？答：此如絲竹之聲亦生名解，似而非實。

Regarding the verse commentary stating: "The Vaibhāṣikas [maintain that] ... because it can be understood,"[the opponent] is confused about the Sarvāstivāda tenet. The Nyāyānusāra states:
"The Vibhāṣā says that name, sentence, and phoneme each have three kinds. The three kinds of name are: name, name-group (nāmakāya), and many-name-group (pada-nāmakāya); sentence and phoneme are likewise. Names have many grades: namely, produced from one syllable (akṣara), produced from two syllables, or produced from many syllables. As for those produced from one syllable: when pronouncing one syllable, there can only be a name; when pronouncing two syllables, it is called a name-group. Or some say: when pronouncing three syllables, it is called a many-name-group. Or some say: only when pronouncing four syllables is it called a many-name-group. As for those produced from two syllables: when pronouncing two syllables, there can only be a name; when pronouncing four syllables, it is called a name-group. Or some say: when pronouncing six syllables, it is called a many-name-group. Or some say: only when pronouncing eight syllables is it called a many-name-group. Among those produced from many syllables, as for those produced from three syllables: when pronouncing three syllables, there can only be a name; when pronouncing six syllables, it is called a name-group. Or some say: when pronouncing nine syllables, it is called a many-name-group. Or some say: only when pronouncing twelve syllables is it called a many-name-group. Taking this as the gateway, the name-groups and many-groups produced from other multiple syllables should be stated in accordance with reason. Sentences also have many grades: namely, the middle sentence, initial sentence and final sentence, short sentence and long sentence. If produced from eight syllables, it is called a middle sentence; because it is neither long nor short, it is called middle. Thirty-two syllables produce four sentences (pādas); four such sentences form a śloka, and the prose and verse of sūtras and śāstras mostly depend on this number. If produced from six syllables or more, it is called an initial sentence; if produced from twenty-six syllables or fewer, it is called a final sentence. If produced from fewer than six syllables, it is called a short sentence; if exceeding twenty-six syllables, it is called a long sentence. For now, distinguishing the three kinds based on the middle sentence: when pronouncing eight syllables, there can only be a sentence; when pronouncing sixteen syllables, it is called a sentence-group (padakāya). Or some say: when pronouncing twenty-four syllables, it is called a many-sentence-group. Or some say: only when pronouncing thirty-two syllables is it called a many-sentence-group. A phoneme is just a syllable, so it has only one grade: when pronouncing one syllable, there can only be a phoneme; when pronouncing two syllables, it is called a phoneme-group (vyañjanakāya). Or some say: when pronouncing three syllables, it is called a many-phoneme-group. Or some say: only when pronouncing four syllables is it called a many-phoneme-group. By this reasoning, one should state: when pronouncing one syllable, there is a name, no name-group, and no many-name-group; there is no sentence, no sentence-group, and no many-sentence-group; there is a phoneme, no phoneme-group, and no many-phoneme-group. When pronouncing two syllables, there is a name, a name-group, and no many-name-group; there are neither of the three kinds of sentence; there is a phoneme, a phoneme-group, and no many-phoneme-group. When pronouncing four syllables, there are the three kinds of name, etc.; neither of the three kinds of sentence; and the three kinds of phoneme, etc. When pronouncing eight syllables, there are the three kinds of name, etc.; a sentence, no sentence-group, and no many-sentence-group; and the three kinds of phoneme, etc. When pronouncing sixteen syllables, there are the three kinds of name, etc.; a sentence, a sentence-group, and no many-sentence-group; and the three kinds of phoneme, etc. When pronouncing thirty-two syllables, name, sentence, and phoneme each possess all three kinds. Taking this as the gateway, the rest should be stated in accordance with reason."
According to the above treatise text, a sound across multiple moments produces one phoneme; the sound of a preceding moment does not have a phoneme. When one phoneme produces one name and a one-syllable sentence, then three dharmas arise simultaneously from a single moment of sound: namely, phoneme, name, and sentence. If it is a name of two or more syllables, in the preceding moment's phoneme there is not yet a name or sentence; only when reaching the later syllable to complete the grade of name and sentence are there then a name and sentence. In cases where many phonemes produce one name, and many names form one sentence, it is not that every single moment of sound has a phoneme, every single phoneme has a name, and every single name has a sentence; rather, there is a name produced from one syllable, and there can be one syllable, one name, and one sentence, or there may be no one-syllable sentence, as a sentence must have two syllables. For those names produced from two or more syllables, in each case only upon the second syllable, etc., is a two-syllable name produced. If a name is produced from two or up to many syllables, understanding can arise already upon the preceding syllable. This is because the listener has previously habituated the understood meaning, not that in the preceding position a two-syllable name, etc., had already been produced previously. For instance, when uttering the two-syllable name "all conditioned phenomena" (zhuxing / sarvasaṃskāra), therein are two aspects: first, two one-syllable names; second, one two-syllable name. The rest follow this by analogy. Why is this so? Just as when uttering a two-syllable name, one person hears zhu ("all") and one person hears xing ("conditioned phenomena"), each generating one understanding. If hearing the first implies the second, this is due to habituation. If one explains it in this way, it accords well with the minimum requirement for a name, namely, one syllable. Furthermore, as in the sentence "all conditioned phenomena are impermanent" (sarvasaṃskārā anityāḥ), etc.: "all conditioned phenomena" is the intrinsic nature (svabhāva), "impermanent" is the distinguishing attribute (viśeṣa); a name expresses intrinsic nature, a sentence expresses a distinguishing attribute. As for the two syllables "impermanent" (wuchang / anitya), whether a name arises or does not arise is undetermined in meaning. For instance, if one says "impermanence is despicable," impermanence is the intrinsic nature, and despicable is the distinguishing attribute. Logically, these two syllables "impermanent" should have a name; if it functions as a distinguishing attribute, then no name arises. To state it in accordance with truth: whether it functions as intrinsic nature or as a distinguishing attribute, all have a name. For instance, when someone utters the sentence "all conditioned phenomena are impermanent," someone hears only the syllable wu, or hears only the syllable chang, or hears the two syllables wuchang, or hears all four syllables completely. If one hears only the syllable wu, one thinks it merely expresses non-existence. If one hears the syllable chang, one thinks it merely expresses permanence. If one hears them combined, one thinks it expresses impermanence. Yet this impermanence does not yet belong to any dharma; therefore we know that what was heard above does not fully comprehend the sentence. If one hears the utterance "all conditioned phenomena are impermanent," only then does one comprehend the sentence, knowing that this impermanence is the distinguishing attribute of all conditioned phenomena. From this we know by analogy that alongside the four syllables "all conditioned phenomena are impermanent," each has a one-syllable name. Therein, if reaching two syllables, there is also a two-syllable name. Reaching the third syllable, there are one-syllable names and a three-syllable sentence. For instance, if someone merely hears the three syllables "all conditioned phenomena are not," this is taken to express that all conditioned phenomena are non-existent: "all conditioned phenomena" being the intrinsic nature, "non-existent" being the distinguishing attribute. Reaching the fourth syllable, there are one-syllable names, two-syllable names, three-syllable names, as well as three-syllable and four-syllable sentences. Addition and subtraction therein should be contemplated in accordance with reason. Another explanation of why four-syllable names, etc., contain one-syllable names is that the mind of the remote preparation (dūra-prayoga) that generates speech mostly takes these as its object. For instance, when uttering the two syllables "all conditioned phenomena," one must first take as object the distinct individual names expressing "all" and expressing "conditioned phenomena" before one can combine them to form a two-syllable name. The Nyāyānusāra and Vibhāṣā say that name, sentence, and phoneme each have three kinds: first, name; second, name-group; third, many-name-group; sentence and phoneme are likewise. Names have many grades: produced from one syllable, produced from two syllables, and produced from many syllables. In the Vibhāṣā, names have six grades: one syllable, two syllables, four syllables, eight syllables, sixteen syllables, and thirty-two syllables. Based on this treatise, it states this according to the doubling of grades and the limit of one verse. However, strictly speaking, it is not that there are no names produced from three syllables, five syllables, seven syllables, nine syllables, etc. The treatise cited above states: names produced from one syllable, names produced from two syllables, and names produced from many syllables. Regarding names produced from many syllables: those produced from three or more syllables are all multi-syllable names. This is because according to the grammar of the western region, expressing two is called a group (kāya) due to having the meaning of aggregation, and expressing three or more is always called "many"; or it is said that four syllables or more are called produced from many syllables, because it doubles two syllables. For those produced from one syllable: when pronouncing one syllable it is merely called a name, and when pronouncing two syllables there is a name-group. Regarding "many-name-group" there are two explanations: or some say, when pronouncing three syllables it is called a many-name-group; or some say, when pronouncing four syllables it is called a many-name-group. Someone asks: "Concerning the 'many-name-group,' one says that pronouncing three syllables is called a many-name-group, while the other says that pronouncing four syllables is called a many-name-group. What is their meaning?" The explanation says: "As for three syllables producing a name being called a many-name-group: the first and second syllables form one group, the first and third syllables form another group, and the second and third syllables form another group. According to the rules of grammar (śabdavidyā), only when three or more are spoken can it be called 'many'. If four syllables producing a name is called a many-name-group, then contiguous combinations form three groups, which is called a many-name-group." Another explanation: "It is called a many-name-group merely with respect to the plurality of names, not with respect to multiple groups." Another explanation: "A group added onto a group is called a many-name-group." According to the above three explanations, they contradict one another; they merely recite them all without determining what is right or wrong. Why? Based on what the treatise explains, it only says that name, sentence, and phoneme each have three grades: namely, one, two, and many. One does not form a group because the name is single; two forms a group because the names are two; and the third should also be called a many-name-group because the names are many. Why would the grades of one and two be the one and two of names, while for the third grade of 'many', its count would suddenly be explained as groups? If here it were called a many-name-group because the groups are many, that would be the third explaining the number of groups, in which case one should also speak of group and group-group. According to the second explanation, four syllables producing 'many' by contiguous pairing should be called a group-group. Furthermore, having themselves cited grammar to the effect that three or more can be called 'many', why do they again say that a group added onto a group is called a many-group? If so, those produced from three syllables would also have two groups when combined in sequence, which would be a group added onto a group, and should also be called a many-name-group. Although one explanation takes it merely with respect to the plurality of names as a many-name-group, since it misses the rationale, it is like a worm eating wood [and accidentally forming characters]. What is explained here as one, two, and many are all stated with respect to names, not discussed with respect to groups. The Vibhāṣā has further explicit text, which should be checked and recorded. Furthermore, quoting the text as saying that three syllables producing a name is called a many-name-group is an error, because three syllables producing a name is only a single three-syllable name. Here the two explanations differ: one is based on becoming 'many' being called a many-group, meaning three or more; the other is based on the grade doubling, meaning four or more names doubling the previous. Following both explanations, neither contradicts the meaning, which is why the treatise presents both. However, there are only the entities of phoneme, name, and sentence, but no entity of verse. This is because regarding phoneme, name, and sentence, across the five destinies and various lands, whether dull or wise, all who can speak produce phonemes through sound; relying on phonemes there are names, and relying on names there are sentences; because all can express intrinsic nature and distinguishing attributes, all cause the comprehension of intrinsic nature and distinguishing attributes. Therefore, verses can only be composed by those who are literate; they are merely the arrangement of phonemes and even names and sentences, without any separate entity. Question: If one utters a single name and a single syllable, and with respect to names of various regions different understandings are produced, are there many names or is there one name? Is there one syllable or are there many syllables? Answer: Syllables are only one, universally applicable across various regions without difference. Names are not the same: if it is speech of this region, there is only the name of this region, and no names of various [other] regions. How is this known to be so? If it were not so, then uttering only a name produced from one syllable should have many names. Since the treatise does not speak of many names, we know there is only one name. Furthermore, when generating speech, the mind of the remote expedient means takes as object only the name of one region. Question: If so, why do multiple understandings arise? Answer: This is like the sounds of strings and flutes also producing the understanding of names: it resembles them, but is not real.